Colorectal cancer (CRC) is the second leading cause of cancer death in the United States. Higher circulating concentrations of 25-hydroxyvitamin D3 (25[OH]D3), the best marker of total vitamin D exposure, may lower the risk of developing and dying from CRC by over 50%. Currently, there are no accepted modifiable (?treatable?) biomarkers of risk for CRC (analogous to lipid biomarkers for heart disease). However, results from our previous clinical trials indicate that vitamin D supplementation can favorably modify the expression of multiple biomarkers of risk for CRC (e.g., p21, bax) in the normal colorectal mucosa. The expression of cyclooxygenase-2 (COX-2) and 15-prostaglandin dehydrogenase (15-PGDH) are key inflammation-related biomarkers of CRC risk that are favorably modified by vitamin D in cancer cell lines, but the effects of vitamin D supplementation on COX-2 and 15-PGDH in humans are unknown. Additionally, results from our recent study indicate that the association of 25(OH)D3 with lower risk of colorectal adenoma (the immediate precursor to most CRCs) differs by common vitamin D binding protein (DBP) isoforms (determined by functional, genetic polymorphisms) that affect circulating DBP and 25(OH)D3 concentrations and vitamin D metabolism. The goals of this project are to investigate the unknown: 1) effects of vitamin D on COX-2 and 15-PGDH expression in the normal colorectal mucosa, and 2) interactions between 25(OH)D3 and vitamin D binding protein (DBP) genotypes/isoforms in relation to CRC risk and survival. We hypothesize that: 1) vitamin D supplementation favorably modifies COX-2 and 15-PGDH expression in the normal-appearing rectal mucosa of colorectal adenoma patients, and 2) the associations of circulating 25(OH)D3 concentrations with CRC risk and survival differ by DBP isoform. For Aim 1, we will measure COX-2 and 15-PDGH expression in biopsies of the normal-appearing rectal mucosa of colorectal adenoma patients (n = 104) in a randomized clinical trial, using automated immunohistochemistry and quantitative image analysis. We will estimate the effects of vitamin D supplementation (1,000 I.U./day) over one year on the two biomarkers using mixed linear regression models. For Aim 2, we will pool previously collected questionnaire, vitamin D assay, and genotyping data from two large prospective cohort studies to investigate whether DBP isoforms modify the association of 25(OH)D3 concentrations with CRC risk in a nested case-control study (n = 1,327 incident CRC cases, 979 matched controls) using multivariable logistic regression, and with CRC survival in a prospective cohort study (n = 1,327 total CRC cases; 479 CRC-specific deaths) using Cox-proportional hazards regression. This research will help elucidate the anti-neoplastic effects of vitamin D, the development of ?treatable? biomarkers of risk for CRC, and the development of ?personalized? vitamin D recommendations, based on one?s inherited genotypes, ultimately reducing CRC incidence and mortality.